A one-time settlement offered too generously recovers less than the borrower would have paid under a restructured schedule. A settlement offered too conservatively is rejected — and the institution proceeds to SARFAESI, spending 18 months and significant legal costs to recover less than the settlement would have produced. The Hard Bucket Agent AI calculates the settlement range that maximises recovery while remaining within the borrower's demonstrated capacity to accept — and does so before the negotiation begins.
Why settlement calibration is a financial modelling problem, not a negotiation instinct
The decision to offer a one-time settlement, and what discount to offer, is one of the most financially consequential decisions in NPA management. The institution is effectively choosing between two recovery pathways: a discounted immediate payment versus a legal recovery process that will take 12–36 months, incur significant costs, and recover an uncertain amount that depends on asset valuation at auction, legal outcome, and market conditions for the asset type.
Most institutions approach this decision with a combination of gut instinct, precedent-based rules (offer 70% of principal), and negotiation — and produce inconsistent outcomes that often over-discount strong recovery cases and under-discount weak ones. The Hard Bucket Agent AI models both recovery pathways explicitly for each account and derives the settlement range that is more attractive to the institution than the legal pathway, while remaining within the range the borrower is financially capable of accepting.
The settlement model: what the AI calculates
Institution minimum
Above NPV of legal recovery
74% of outstanding
Within borrower capacity
If borrower has liquidity
From property sale or family
The two pathways — why the model compares them
The settlement calibration model compares two recovery pathways for every account. The legal recovery pathway — SARFAESI possession, auction, DRT proceedings — produces a net present value of ₹28.4 lakhs in this case: the estimated ₹33.8 lakh auction recovery (65% of market value, based on the 3-year average recovery ratio for residential properties in Bengaluru) less ₹2.2 lakhs in legal and possession costs, discounted back 22 months at the institution's cost of capital.
The settlement pathway — a one-time payment of ₹34.5 lakhs received within 30 days of the offer being accepted — produces an immediate cash recovery that is ₹6.1 lakhs above the NPV of the legal pathway, without the 22 months of management time, legal costs, and auction uncertainty. The settlement is the superior option for the institution — not as a concession to the borrower, but as the financially rational choice given the alternatives.
This is the insight that the Hard Bucket AI makes explicit. Settlement negotiation without an explicit NPV comparison to the legal alternative typically results in the institution either offering too much (because the legal pathway looks expensive) or too little (because the outstanding amount anchors the negotiation). The model removes the anchoring bias and replaces it with a rational floor: the institution should accept any settlement above the NPV of the legal recovery pathway.
What the board-approved settlement authority framework looks like
| DPD Range | Max Waiver (% of outstanding) | Approval Authority | Settlement Payment Terms | Board Reporting |
|---|---|---|---|---|
| DPD 90–120 (early NPA) | Up to 10% | Collections Head | Full settlement within 30 days | Quarterly board report |
| DPD 121–180 | Up to 20% | Credit Risk Head + CCO | Full settlement within 45 days | Quarterly board report |
| DPD 181–360 | Up to 35% | MD + Credit Committee | Up to 3 instalments over 90 days | Board approval per case above ₹25L |
| DPD 360+ (chronic NPA) | Up to 55% | Board Risk Committee | Up to 6 instalments over 180 days | Board approval required · All cases |
| Write-off consideration | Over 55% | Board Resolution required | Any amount — NPV positive vs legal | Board resolution + RBI reporting |
Settlement is not a concession — it is a recovery strategy with a financial model behind it
The institution that treats settlement as a reluctant compromise — offering a discount only because the legal route seems too hard — will always offer either too much or too little. The institution that treats settlement as a financially modelled recovery pathway — knowing exactly what the legal alternative is worth in NPV terms, knowing the borrower's demonstrated payment capacity, and knowing the market conditions for its security — will consistently make settlement offers that are superior to the legal alternative for the institution, acceptable to the borrower, and justifiable to the board on financial grounds. The Hard Bucket Agent AI builds that financial model for every NPA account before the negotiation begins. Settlement stops being a gut feel and becomes a financial decision.